Seabird predation by white shark and Cape fur seal at Dyer Island

R.L. Johnson1*, A. Venter2, M.N. Bester1 & W.H. Oosthuizen3 1Mammal Research Institute, Department of Zoology & Entomology, University of Pretoria, Pretoria, 0002 South Africa 2Western Cape Nature Conservation Board, Private Bag X13, Hermanus, 7200 South Africa 3Branch Marine and Coastal Management, Department of Environmental Affairs and Tourism, Private Bag X2, Roggebaai, South Africa Received 31 August 2004. Accepted 10 November 2005 Both the white shark (Carcharodon carcharias) and Cape fur seal (Arctocephalus pusillus pusillus) prey on and/or attack seabirds in South Africa. The Dyer Island region abounds in these predators, as well as seabirds, including the African penguin (Spheniscus demersus), (Phalacrocorax capensis), (P. neglectus), white-breasted cormorant (P. carbo), and (P. coronatus). Between August 1999 and February 2001, predatory interactions among these taxa were quantified and qualified by the routine collection and inspection of seabird carcasses and injured birds, as well as opportunistic observations of live attacks. White sharks are infrequent predators of seabirds at Dyer Island, perhaps due to an abundance of Cape fur seals (a preferred prey), anti- predator behaviour by penguins, and seabirds not being a preferred prey type. Cape fur seals were more conspicuous seabird predators, each year attacking adult penguins (1.99–2.52%), white-breasted cormorants (5.21–5.72%), and crowned cormorants (3.13%), as well as a single bank cormorant. Cape fur seals killed an estimated 0.83–1.09% of the fledgling Cape cormorants. Attacks on penguins at the island are crepuscular as birds depart to and return from foraging grounds. Fledgling Cape cormorants are frequently attacked when alighting on water following disturbance and failed flight attempts, or for bathing. Human disturbance, which may force birds to take to the water, may cause inter-annual differences in the predation impact on this taxon. Key words: African penguin, Cape fur seal, cormorants, Dyer Island, predation, white shark.

INTRODUCTION southern Africa including the Cape fur seal Many South African seabird populations are (Shaughnessy 1978; Broni 1985; Marks et al. declining due to various threats (Barnes 2000). 1997; David et al. 2003), the white shark (Randall Both human-induced factors such as oil pollution, et al. 1988) and the killer whale (Orcinus orca) egg harvesting, human disturbance (Hockey & (Randall & Randall 1990). The Cape fur seal is a Hallinan 1981; Shelton et al. 1984; Crawford et al. most conspicuous seabird predator and has been 2000) and natural effects, including predation, recorded hunting and feeding on gannets, Morus interspecific competition for nesting sites, modifi- capensis (Du Toit 2002; David et al. 2003), cormo- cation to fish stocks (Crawford et al. 1989, 1990; rants, Phalacrocorax spp. (Marks et al. 1997) and Crawford & Dyer 1995; Marks et al. 1997; David penguins (Shaughnessy 1978; Crawford et al. et al. 2003) have been implicated as possible 2001; Du Toit 2002). Up to 7.1% of the fledgling contributors to seabird declines. Quantitative population of Cape cormorants at Dyer Island may accounts of seabird predation are rare (e.g.Du Toit fall victim to seals annually (Marks et al. 1997), et al. 2004), and therefore the relative importance while Du Toit (2002) estimated that 0.9% of the of predatory attacks in population declines is African penguin population on Ichaboe Island, currently unknown. , succumbed to seal predation. David A number of apex predators attack, feed on or et al. (2003) calculated that 7.4% of the fledgling induce anti-predator behaviour in seabirds in population at Malgas Island, South *E-mail: [email protected] Africa, was killed by Cape fur seals during the South African Journal of Wildlife Research 36(1): 00–00 (April 2006) 2 South African Journal of Wildlife Research Vol. 36, No. 1, April 2006

Fig. 1. Map of the Dyer Island study site with zonation illustrated. Also indicated is the location of three attempted attacks by white sharks on kelp gulls.

November 2000 – March 2001 breeding season. classed as ‘Endangered’ due to ongoing popula- These estimates are large enough to warrant more tion declines (Crawford 2005c). The Cape cormo- attention to such attacks. rant is classed as ‘Near-threatened’ due to a Southern Africa is one of the centres of abundance population decrease from 277 000 pairs to for the white shark (Compagno 1991) where it c. 72 000 pairs between 1977 and 1996 (Barnes ranges from southern Mozambique to Namibia. 2000). The white-breasted cormorant is not White sharks were responsible for a majority of the threatened in South Africa (Barnes 2000). injured and dead African penguins recovered from Most research efforts have been directed at the shoreline of St Croix Island and Bird Island, single predator-prey interactions, despite the need Algoa Bay, South Africa (Randall et al. 1988). for ecosystem based management (David et. al. However, only a single African penguin has been 2003). The diversity and abundance of predator recovered from a white shark stomach (Bass et al. and prey species occurring within the Dyer Island 1975). area enable an unprecedented array of interspecific Resident birds on Dyer Island identified as predatory interactions to be documented and vulnerable to attack by either the Cape fur seal investigated. The aim of this paper is to determine or white shark include the African penguin the comparative importance of seabird predation (Shaughnessy 1978; Randall et al. 1988) and the by white sharks and Cape fur seals on the resident four cormorant species (Marks et al. 1997; Du Toit seabird populations at Dyer Island. 2002; David et. al. 2003; Crawford 2005a–d). The African penguin is the only endemic penguin in STUDY AREA southern Africa, and is listed as ‘Vulnerable’ by the The Dyer Island complex (34°41’S; 19°25’E) lies World Conservation Union (IUCN) due to a persis- off the Western Cape Province of South Africa, tent population decline (Crawford et al. 1990; and consists of two islands (Fig. 1). Dyer Island is Barnes 2000). The IUCN further classifies the the largest with a surface area of 20 ha and is ex- endemic crowned cormorant as ‘Near-threatened’ tensively inhabited by seabirds; Geyser Rock is because of its small population size (Barnes the smaller island, lying 230 m southwest of Dyer 2000). The endemic bank cormorant is currently Island.Geyser Rock is host to an estimated 55 000 Johnson et al.: Seabird predation by white shark and Cape fur seal at Dyer Island 3

Cape fur seals (J.H.M. David, pers. comm.). Dyer 1999 and October 2000, December 1999 and Island was subdivided into six zones to facilitate November 2000, and January 2000 to December the present study (Fig. 1) as outlined below. 2000 were summed independently and averaged to estimate the breeding population size on Dyer METHODS AND MATERIALS Island conservatively. Monthly nest counts of Predatory encounters were investigated through all occupied cormorant nests, defined as sites opportunistic observation of attacks, collection of defended by adult birds or sites with eggs or bird carcasses and observation and/or collection chicks, were carried out. Peak nest counts of of injured seabirds between August 1999 and the Cape, bank, crowned and white-breasted February 2001. Daily searches for injured birds cormorant species were multiplied by two (breed- and washed-up seabird carcasses in the intertidal ing pairs) to estimate the numbers of cormorants zone were carried out during a circumperambulation at Dyer Island conservatively. of Dyer Island. All seabird carcasses discovered The predatory impact of Cape fur seals on the were removed, examined and buried. The taxa African penguin population was taken as the (species level), date, location (zone), age class, percentage of the adult population (established by sex (for sexually dimorphic species only) and moult counts) killed between January 2000 and stage of decomposition (fresh vs skeletal remains) December 2000. The fledgling population of Cape were recorded postmortem. Where applicable, cormorants was calculated by multiplying the peak injuries on the carcasses were described, and nest count by 2.36 (average eggs/nest) (Crawford when possible the predator identified. Depending 1992). The survival coefficients between eggs and on the severity of the injury, and the degree of chicks (0.87) and chicks to fledglings (0.91) were disturbance the collection of data would cause, then used to establish the fledgling population injured seabirds were both collected for examina- (Crawford 2005d). The annual impact of predators tion and treated, or simply examined in the field on other cormorant species was taken as the and released. During examination the wound percentage of the population (established by peak characteristics were described, and when possi- nest count) killed during the corresponding 12 ble the predator identified. Each victim’s location months (January–December 2000). For each (zone), age-class, sex and the wound sever- analysis, the lower minimum estimate includes ity (superficial, serious, mortal) were routinely only carcasses of birds definitely killed by Cape fur recorded. seals, while the upper minimum includes skeletal Live attacks were observed opportunistically remains that could have come from seabirds killed from various vantage points, including an observa- by fur seals. Both estimates are conservative, as tion tower (height 4 m), within the living compound, unknown proportions of carcasses do not wash up during daily circum-island research patrols, and on Dyer Island. Diel and seasonal trends in attack from a research vessel anchored at various frequency were tested using analysis of frequency locations in the near vicinity of the island (Fig. 1). tests (χ2 test), with significance set at the 5% level. Predation cues included: splashing; hovering kelp gulls (Larus dominicanus); a splashing seal, or RESULTS the formation of an oil slick as a result of natural internal oils emanating from a disembowelled Seabird carcasses seabird. The predator and prey species involved Between August 1999 and January 2001 a total (cormorants to genus level only) were recorded, of 812 seabird carcasses were collected (Table 1). together with the time, location of the event, local Of these the African penguin, Cape cormorant, environmental conditions, and a description of the bank cormorant, crowned cormorant, white- encounter. breasted cormorant and Cape gannet showed Moulting and breeding population trends in the evidence of predation.Predation by Cape fur seals African penguin and four cormorant species were evidently caused the death of 71.6% of the birds, established by fortnightly counts of moulting while no carcasses displayed evidence of a white penguins, and a monthly nest count of all birds shark bite. Approximately 7.5% of the carcasses (Crawford & Boonstra 1993). A total of 34 counts showed no evidence of predation, while advanced of moulting African penguins were completed decomposition prevented the establishment of the between August 1999 and January 2001. Moult cause of death in the remaining 20.9% of car- counts (adults only) conducted between November casses.Some 87.7% of the corpses that displayed 4 South African Journal of Wildlife Research Vol. 36, No. 1, April 2006

Table 1. Cause of death in the various species of seabird carcasses collected from the intertidal region of Dyer Island between September 1999 and January 2001.

Species Cause unknown No. injuries C.F.S. wounds W.S. wounds Total no. of apparent present present carcasses

African penguin 45 (21.0%) 15 (07.0%) 154 (72.0%) 0 (00.0%) 214 Cape cormorant 115 (17.4%) 42 (11.3%) 396 (71.6%) 0 (00.0%) 553 Bank cormorant 0 (00.0%) 0 (00.0%) 1 (100%) 0 (00.0%) 1 Crowned cormorant 1 (07.7%) 3 (23.1%) 9 (69.2%) 0 (00.0%) 13 White-breasted cormorant 3 (12.5%) 1 (04.2%) 20 (83.3%) 0 (00.0%) 24 Gannet 0 (00.0%) 0 (00.0%) 1 (100%) 0 (00.0%) 1 Cormorant sp. 6 (100%) 0 (00.0%) 0 (00.0%) 0 (00.0%) 6

C.F.S. = Cape fur seal, W.S. = white shark.

Fig. 2. White shark-inflicted wounds on African penguins; a, a single elongated puncture is visible on the dorsal surface; b, multiple punctures to the body cavity characterize this bite.

Cape fur seal inflicted wounds were ‘degloved’ as identified (Table 2). Additionally, three encounters defined by Marks et al. (1997). Abdomen and neck between white sharks and kelp gulls were observed bites accounted for the remaining 5.2% and 7.1% (Table 2). White sharks attacked floating gulls of injury to carcasses, respectively. twice within the channel area, adjacent to Geyser Rock (attacks B and C, Fig. 1). During attack ‘B’ Injured seabirds the white shark successfully hit, and killed, the gull Injured seabirds included the African penguin and Cape cormorant. Excluding birds contami- Table 2. Live attacks by various predators on seabirds nated by oil, 26 of 31 birds were African penguins, occurring at Dyer Island. ‘Unsuccessful attempt’ repre- while the remainder were Cape cormorants. sents all occasions when predator(s) made an unambig- uous attempt to attack potential prey but failed. Predation was adjudged the cause of injury in only 10 birds, with three Cape cormorants and five White shark Cape fur seal penguins sporting wounds inflicted by Cape fur seals. Two penguins showed evidence of white Successful attacks shark bites, the only evidence of white sharks African penguin 0 92 attacking African penguins in this study (Fig. 2a,b). Cormorant sp. 0 79 Kelp gull 1 0 Unknown 0 32 Live attacks A total of 204 attacks by Cape fur seals on Unsuccessful attacks African penguins and cormorant spp. were African penguin 0 0 Cormorant sp. 0 1 observed. Of these 46.6% involved the African Kelp gull 2 0 penguin, 38.7% a cormorant sp., and in the Unknown 0 0 remaining 14.7% of attacks the victim was not Johnson et al.: Seabird predation by white shark and Cape fur seal at Dyer Island 5

Fig. 3. Seasonal trends in seabird predation at Dyer Island; a, collected seabird carcasses; b, observed attacks by Cape fur seals. but did not return to feed on it. The third encounter and late (January to March) summer periods χ2 (attack A) was a breach from a white shark of ( (0.05,3) = 44.2, P < 0.01) (Fig. 3a). Predation on c. 380 cm total length towards two hovering gulls Cape cormorants by Cape fur seals showed a sig- χ2 that failed to make contact. nificant seasonal pattern ( (0.05,3) = 232.3, P < 0.001) with large numbers of carcasses collected Seasonal trends in Cape fur seal attacks in the early and late summer periods in both 1999 Attacks on African penguins by Cape fur seals and 2000 (Fig. 3a). White-breasted cormorant car- χ2 were seasonal, with few penguin carcasses casses appeared aseasonally ( (0.05,3) = 1.87, P > collected during the early (October to December) 0.50) (Fig. 4a). The recovery of crowned cormo-

Fig. 4. Seasonal patterns of carcass collection for three species of cormorants resident on Dyer Island; a, white- breasted cormorant; b, crowned cormorant; c, bank cormorant. 6 South African Journal of Wildlife Research Vol. 36, No. 1, April 2006

Fig. 5. Nesting and moulting trends of seabirds at Dyer Island; a, African penguins; b, Cape cormorants.

χ2 χ2 rant carcasses was seasonal ( (0.05,3) = 8.4, P < 14 contingency table, (0.005,13) = 45.25, P < 0.01). 0.05) with 92% of the carcasses collected in the Crepuscular peaks in penguin attacks were evi- early (April to June) and late (July to September) dent, while attacks on cormorants increased winter period (Fig. 4b). Only one carcass of a bank throughout the morning period, and peaked at cormorant was recovered in October 1999 around 13:00, before trailing off (Fig. 6b). (Fig. 4c). The seasonal trends in attacks by Cape fur seals on seabirds mimicked trends of carcass Population trends of seabirds recovery (Fig. 3b). Live attacks on African penguins The African penguin is a seasonal breeder at were seasonal, with few attacks seen during the Dyer Island, with peak nest counts recorded χ2 late summer period ( (0.05,3) = 23.3, P < 0.001). The between March and August (Fig. 5a), while the majority of the attacks observed involving cormo- peak moulting period is between October and χ2 rant spp. occurred in early summer ( (0.05,3) = 27.3, December (Fig. 5a). Nest counts of Cape cormo- P < 0.001) (Fig. 3b). rants showed that in 1999 nesting began in earnest during September, while in 2000 it began slightly Population trends of seabirds earlier in August (Fig. 5b). The African penguin is a seasonal breeder at Dyer Island, with peak nest counts recorded Importance of predation in seabird mortality between March and August (Fig. 5a), while the The African penguin breeding population was peak moulting period is between October and estimated at 5081 birds. During the corresponding December (Fig. 5a). Nest counts of Cape cormo- 12 months (Jan 2000 – Dec 2000) a total of 134 rants showed that in 1999 nesting began in earnest penguin carcasses were collected of which 101 during September, while in 2000 it began slightly showed evidence of Cape fur seal attack. The earlier in August (Fig. 5b). remaining 33 birds were either too decomposed for cause of death to be established (27 birds), or Diurnal trends in Cape fur seal attacks no injuries were apparent (six birds). Therefore, a Significant diurnal trends in the minimum rate of lower minimum of 1.99% (and an upper minimum χ2 live attacks existed for both penguins ( (0.005,13) = of 2.52%) of the breeding population died from χ2 71.13, P < 0.01) and cormorants ( (0.005,13) = 24.21, Cape fur seal attacks during this period. White- P < 0.05) (Fig. 6).Differences existed in the diurnal breasted cormorants were the most vulnerable patterns of attack on penguin and cormorants (2 × cormorant species to seal attack, which resulted in Johnson et al.: Seabird predation by white shark and Cape fur seal at Dyer Island 7

Fig. 6. Diel patterns in live attacks on seabirds at Dyer Island.

Table 3. Predatory impact of Cape fur seals on cormorant spp. at Dyer Island between January 2000 and December 2000.

Species Peak nest count Population estimate Seal predation Annual predation impact (%)

Bank cormorant 42 84 Min: 0 0.00 Max: 0 0.00 Crowned cormorant 128 256 Min: 8 3.13 Max: 8 3.13 White-breasted cormorant 96 192 Min: 10 5.21 Max: 11 5.72 the death of at least 5.72% of the population observation used to investigate the trends in (Table 3). Crowned cormorants (3.13%) appear attacks on seabirds (Randall et al. 1988). The less affected and bank cormorants virtually unaf- most conspicuous difference between the two fected (Table 3). The peak nest count of Cape ecosystems is the presence of a large Cape fur cormorants (18 105) translates to a fledgling popu- seal colony at Dyer Island, but not at at St Croix lation of 33 827 fledglings during the 2000/ Island. This may have a number of direct and 2001-year. A lower minimum of 280 and an upper indirect effects on the behaviour of penguins and minimum of 370 fledgling Cape cormorants were white sharks. killed by Cape fur seals during the corresponding Pinnipeds represent a primary prey for white period (Janoury 2000 – December 2000). This sharks (Le Boeuf 1982; Klimley et al. 1992; suggests an annual minimum predation rate of Compagno et al. 1997). The c. 55 000 strong seal between 0.83% and 1.09% of the fledgling popula- colony may direct the attention of white sharks tion. towards this prey, and consequently, less effort may be spent investigating alternative prey that is DISCUSSION not as energetically rewarding, or not part of their White sharks do not appear to be a major threat to typical diet (Bass et al. 1975; Cliff et al. 1989, seabirds at Dyer Island, in contrast to the situation 1996). Anti-predator behaviour (fleeing to and at St Croix Island, despite similar methods of from an island) of African penguins in the vicinity of 8 South African Journal of Wildlife Research Vol. 36, No. 1, April 2006 pinnipeds, and other predators, has been well nesting began. Cape cormorants tended to be documented (Randall & Randall 1990), and was attacked throughout the day, with a peak around frequently observed at Dyer Island (this study). midday. Disturbance, failed flight attempts and Such behaviour would also make penguins less bathing often resulted in large numbers of fledgling vulnerable to investigatory bites from white cormorants swimming in the waters adjacent to sharks. Dyer Island. In particular when approach from Injured survivors offered the only evidence of humans causes a chain reaction of panic, masses white sharks attacking penguins. The relatively of fledglings flee to the water. The conservative gentle nature of these bites suggests curiosity or estimate that 1.09% of the Cape cormorant fledg- inspection as possible motivation, rather than ling population succumb to seal predation is feeding.This is consistent with the wide diversity of noticeably lower than the 7.1% calculated by non-prey items that have been bitten by white Marks et al. (1997) for Dyer Island. One factor that sharks (Collier et al. 1996), as well as, infrequent may have contributed to the large difference was consumption of non-prey species such as humans the level of human disturbance experienced by the (Homo sapiens) (Burgess & Callahan 1996), sea respective fledgling populations. Extensive obser- otters (Enhydra lutris) (Ames & Morejohn 1980) vation periods, from a tower constructed on the and African penguins (Bass et al. 1975) following southwestern point of Dyer Island, were necessary an attack. for land based observations of the channel area Penguins were attacked by Cape fur seals during the study of Marks et al. (1997). This may seasonally with lowest attacks during November have contributed to more fledglings entering the and December, the peak moulting period of the water due to human disturbance by observers penguins. Penguins remain on land during their moving to and from the tower, and movement on moult, except for preening and drinking (Cooper the tower, than during the present study. 1974) and restrict foraging to a minimum (Broni It can be concluded that the continual decrease 1985). Penguins at Dyer Island appear to be in penguin numbers in South Africa requires most vulnerable to seal predation during foraging management to attempt to minimize all factors related movements away from, and subsequent contributing to the decline, including seal predation return to, the island. The crepuscular pattern of (David et al. 2003). Active removal of rogue seals such live attacks (this study) coincides with the at Malgas Island saw significant declines in attack mean departure (08:00) and arrival time (17:30) of rates in 1999 and 2000 (David et al. 2003). Distur- foraging penguins during the breeding season at bance, human or otherwise, of penguins on the Marcus Island, Western Cape (Wilson et al. 1988). island does not appear to play a major role in Cape fur seals take at least 1.99% to 2.52% of the increasing seal predation rate, as most penguins adult breeding population of penguins at Dyer are attacked when travelling to and from their feed- Island, an underestimate as only some carcasses ing grounds. This quantitative account of Cape fur end up on Dyer Island’s shore. In addition, the loss seals attacking and consuming bank, crowned of a breeding adult results in the loss of its clutch, and white-breasted cormorants in southern Africa thus further compounding the predation impact. shows that these attacks are comparatively infre- Cape cormorants appear vulnerable to predation quent and opportunistic. However, the impact on by seals due to their frequent landing in the waters the small local cormorant populations is significant surrounding Dyer Island. The presence of cormo- due to their low abundance and vulnerability to rants in the inshore area of Dyer Island is not stochastic disturbances. Human disturbance in necessarily related to feeding, as feeding occur up particular, should be minimized in areas and at to 40 km from shore (Berry 1976) in sub-surface times when cormorants are breeding and fledg- waters (Crawford 2005d). The alighting of cormo- lings are present. rants on waters close to shore followed failed flight attempts and/or onshore disturbance (this study) ACKNOWLEDGEMENTS and the seasonal peak in attacks corresponded to This work formed part of a larger study funded by the annual fledging periods of this species. The the Department of Environmental Affairs and period between egg laying and fledging is from 10 Tourism, the University of Pretoria, Western Cape to 12 weeks (Crawford 2005d), and in both 1999 Nature Conservation Board and the International and 2000, attacks on fledgling Cape cormorants Association of Impact Assessment – South African concomitantly rose sharply three months after affiliate. For permission to work on Dyer Island, Johnson et al.: Seabird predation by white shark and Cape fur seal at Dyer Island 9 and the use of facilities we thank the Western predators in the Benguela system – implications of Cape Nature Conservation Board. For logistical their trophic position. S. Afr. J. mar. Sci. 12: 675–687 CRAWFORD, R.J.M. & BOONSTRA, H.G.V.D. 1993. (research vessel) and additional support we thank Counts of moulting and breeding jackass penguins Andre Hartman and Marine Dynamics. (Spheniscus demersus): a comparison at Robben Island, 1988–1993. Mar. Ornithol. 22: 213–219. REFERENCES CRAWFORD, R.J.M. & DYER, B.M. 1995. Responses by AMES, J.A.& MOREJOHN, G.V.1980.Evidence of white four seabird species to a fluctuating availability of shark (Carcharodon carcharias) attacks on sea Cape anchovy (Engraulis capensis) off South Africa. otters (Enhydra lutris).Calif.Fish Game 66:196–209. Ibis 137: 329–339. BARNES, K.N., 2000. The Eskom Red Data Book CRAWFORD, R.J.M., DAVIS, S.A., HARDING, R.T., of Birds of South Africa, Lesotho and Swaziland. JACKSON, L.F., LESHORO, T.M., MEŸER, M.A., BirdLife South Africa, Johannesburg RANDALL, R.M., UNDERHILL, L.G., UPFOLD, L., BASS, A.J., D’AURBREY, J.D. & KITNASAMY, N. 1975. VAN DALSEN, A.P., VAN DER MERWE, E., WHIT- Sharks of the east coast of southern Africa. The TINGTON, P.A., WILLIAMS, A.J. & WOLFAARDT, families Odontaspididae. Scapanorhynchidae, Isuri- A.C. 2000. Initial impact of the Treasure oil spill on dae. Cetorhinidae. Alopiidae and Rhiniodontidae. seabirds off western South Africa. S. Afr. J. mar. Sci. Invest. Rep. Oceanogr. Res. Inst. Durban 39: 1–102. 22: 157–176. BERRY, H.H. 1976. Physiological and behavioural ecol- CRAWFORD, R.J.M., DAVID, J.H.M., SHANNON, L.J., ogy of the Cape Cormorant Phalacrocorax capensis. KEMPER, J., KLAGES, N.T.W., ROUX, J-P., Madoqua 9: 5–55. UNDERHILL, L.G., WARD. V.L., WILLIAMS, A.J. & BRONI, S.C.1985.Social and spatial foraging patterns of WOLFAARDT, A.C. 2001. African penguins as the jackass penguin (Spheniscus demersus).S.Afr.J. predators and prey – coping (or not) with change. Zool. 20: 241–245. S. Afr. J. mar. Sci. 23: 435–447. BURGESS, G.H. & CALLAHAN, M. 1996. Worldwide CRAWFORD, R.J.M. 2005a. The white breasted patterns of white shark attacks on humans. In: A.P. cormorant. In: P.A.R. Hockey et al. (Eds), Roberts’ Klimley & D.G. Ainley (Eds), The biology of Carcha- birds of southern Africa, 7th edn (pp. 575–577). John rodon carcharias (pp. 457–469). Academic press, Voelcker Bird Book Fund, Cape Town. California. CRAWFORD, R.J.M. 2005b. The crowned cormorant. In: CLIFF, G., DUDLEY, S.F.J. & DAVIS, B. 1989. Sharks P.A.R.Hockey et al. (Eds), Roberts’ birds of southern caught in the protective gill nets off Natal, South Africa, 7th edn (pp. 573–574). John Voelcker Bird Africa. 2. The great white shark (Carcharodon Book Fund, Cape Town. carcharias) (Linnaeus).S.Afr.J.mar.Sci.8:131–144. CRAWFORD, R.J.M. 2005c. The bank cormorant. In: CLIFF, G., DUDLEY, S.F.J. & JURY, M.R. 1996. Catches P.A.R.Hockey et al. (Eds), Roberts’ birds of southern of white sharks in KwaZulu-Natal, South Africa and Africa, 7th edn (pp. 577–578). John Voelcker Bird environmental influences. In: A.P. Klimley & D.G. Book Fund, Cape Town. Ainley (Eds), The Biology of Carcharodon carcharias, CRAWFORD, R.J.M. 2005d. The Cape cormorant. In: (pp. 351–362). Academic press, California. P.A.R.Hockey et al. (Eds), Roberts’ birds of southern COLLIER, R.S., MARKS, M.A. & WARNER, R.W. 1996. Africa, 7th edn (pp. 575–577). John Voelcker Bird White shark attacks on inanimate objects along the Book Fund, Cape Town. pacific coast of North America. In: A.P.Klimley & D.G. DAVID, J.H.M., CURY, P., CRAWFORD, R.J.M., Ainley (Eds), The biology of Carcharodon carcharias, RANDALL, R.M., UNDERHILL, L.G. & MEYER, M.A. (pp. 217–222). Academic Press, California. 2003. Assessing conservation priorities in the COMPAGNO, L.J.V.1991. Government protection for the Benguela ecosystem, South Africa: analysing great white shark (Carcharodon carcharias) in South predation by seals on threatened seabirds. Biol. Africa. S. Afr. J. Sci. 87: 284–285. Conserv. 114: 289–292. COMPAGNO, L.J.V.,MARKS, M.A.& FERGUSSON, I.K. DU TOIT, M. 2002. Predatory interactions between Cape 1997. Threatened fishes of the world – Carcharodon fur seals and seabirds at Ichaboa Island, Namibia. carcharias (Linnaeus, 1758) (Lamnidae). Environ. M.Sc. thesis, University of Pretoria, Pretoria. Biol. Fish. 50: 61–62. DU TOIT, M., BARTLETT, P.A.,BESTER, M.N. & ROUX, COOPER, J. 1974. The predators of the jackass penguin J-P.2004. Seal-seabird predation at Ichaboe Island, (Spheniscus demersus). Bull. Brit. Ornith. Club 94: Namibia: prey selection by individual seals. S. Afr. J. 21–24. Wildl. Res. 34: 45–54. CRAWFORD, R.J.M., DAVID, J.H.M., WILLIAMS, A.J. & HOCKEY, P.A.R., HALLINAN, J. 1981 Effect of human DYER, B.M. 1989. Competition for space: recolo- disturbance on the breeding behaviour of jackass nizing seals displace endangered, endemic seabirds penguins Spheniscus demersus. S. Afr. J. Wildl. Res. off Namibia. Biol. Conserv. 48: 59–72. 11: 59–62. CRAWFORD, R.J.M., WILLIAMS, A.J., RANDALL, R.M., KLIMLEY, A.P., ANDERSON, S.D., PYLE, P. & RANDALL, B.M., BERRUTI, A. & ROSS, G.J.B. 1990. HENDERSON, R.P.1992.Spatiotemporal patterns of Recent population trends of the jackass penguin white shark (Carcharodon carcharias) predation at (Spheniscus demersus) off southern Africa. Biol. the South Farallon islands, California. Copeia 3: Conserv. 52: 229–243. 680–690. CRAWFORD, R.J.M., UNDERHILL, L.G., RAUBEN- LE BOEUF, B.J., RIEDMAN, M. & KEYES, R.S. 1982. HEIMER, C.M., DYER, B.M. & MÄRTIN, J. 1992. Top White shark predation on pinnipeds in California 10 South African Journal of Wildlife Research Vol. 36, No. 1, April 2006

coastal waters. Fish. Bull. 80: 891–895. demersus): deductions based on behaviour. Mar. MARKS, M.A., BROOKE, R.K. & GILDENHYS, A.M. Ornithol. 18: 9–12. 1997. Cape fur seal (Arctophalus pusillus pusillus) SHAUGHNESSY, P.D. 1978. Cape fur seals preying on predation on Cape cormorants (Phalacrocorax seabirds. Cormorant 5: 31. capensis) and other birds at Dyer Island, South SHELTON, P.A., CRAWFORD, R.J.M., COOPER, J. & Africa. Mar. Ornithol. 25: 9–12. BROOKE, R.K. 1984. Distribution, population size RANDALL, B.M., RANDELL, R.M. & COMPAGNO, L.J.V. and conservation of the jackass penguin (Spheniscus 1988. Injuries to jackass penguins (Spheniscus demersus). S. Afr. J. mar. Sci. 2: 217–257. demersus) evidence of shark involvement. J. Zool. WILSON, R.P., WILSON, M-P.T. & DUFFY, D.C. 1988. (Lond.) 214: 589–599. Contemporary and historical patterns of African RANDALL, R.M. & RANDALL, B.M. 1990. Cetaceans as penguin Spheniscus demersus: distribution at sea. predators of the jackass penguin (Spheniscus Estuar. Coast. Shelf Sci. 26: 447–458

Corresponding editor: C.D. McQuaid